1. Field of the Invention
The present disclosure generally relates to the field of identifying a molecular substance by a substance's spectra when spectra contains a plurality of colors, and more particularly to identifying an unknown molecular substance by comparing the substance's spectra with a spectra of a known substance.
2. Background
There are two primary categories for the detection of substances by spectral measurement and identification. The first is to detect a sample remotely. Such a system is frequently referred to as performing standoff detection. In this configuration a light source may be used to illuminate the target material from a distance and receive light scattered from the sample in a Lidar type system or it may receive light scattered from natural illumination or it may receive light produced by a self-illuminating material. Remote detection systems currently are Differential Absorption Lidar (DIAL), Differential Scattering (DISC), Infrared Fourier Transform Spectroscopy, (IR FTS), and Raman Lidar. IR FTS can also operate on self luminance or in a passive configuration using natural illumination.
The most common standoff chemical detection techniques are DIAL and DISC. Both systems employ the same basic phenomenology to remotely detect airborne or surface contaminants. In most configurations, such systems transmit a succession of highly monochromatic pulses at and near known absorption lines of materials of interest. These laser pulses reflect off airborne aerosols and/or hard surfaces and are detected by the DIAL or DISC receiver. If the material of interest is not present, the ratio of the reflected signal strength at the absorption wavelength to the reflected signal strength near the absorption wavelength is ideally one. A ratio value of less than one is a measure of the concentration of the material of interest.
Although DIAL and DISC techniques have been shown effective, they possess significant inherent deficiencies. Typically, current systems employ CO2 lasers operating from 9.2 μm to 10.8 μm. CO2 lasers have numerous R- and P-branch transitions with sufficient gain for efficient laser operation. However, switching from one laser transition at one wavelength to another requires retuning the laser. Rapid tuning of the laser cavity requires extremely precise control. Transmitter lasers based upon cavity tuning tend to be complex, require precision instruments to control the tuning, and are prohibitively expensive. Current efforts to transition technology from the former Soviet Union for the frequency conversion of CO2 laser lines using high-pressure ammonia (NH3) is promising, but the technology is not sufficiently mature for deployment. Non-CO2 systems include optical parametric oscillators and optical parametric amplifiers configurations. These also tend to be very complex and, therefore, expensive. The complexity of these devices also reduces their reliability.
A serious limitation of these techniques are that the atmosphere or a potentially contaminated surface cannot be interrogated for multiple chemical species simultaneously with a single laser. A minimum of two pulses must be transmitted for each chemical of interest (one on the absorption line and one adjacent to it). The transmission of multiple pulses is required to evaluate multiple contaminants, even under the best of conditions. In practice, many lines are required to identify just one substance in the presence of interferants. An “interferant” refers to any substance other than the desired substance, which also emits a spectra. The wavelength of these laser pulses must match the absorption line within restrictive tolerances. This complicates system design and system cost, as laser wavelength cannot be selected arbitrarily in most lasers. In this regard, most laser wavelengths are dictated by quantized atomic or molecular energy states. The selected wavelength(s) must be precisely controlled to ensure that they match the absorption line. In addition, many useful detection lines (e.g., 3-5 μm, 8.3 μm) fall outside of the normal CO2 laser gain lines, even for isotopic CO2 lasers. Optical parametric oscillators and amplifier may also be used in such a scenario.
IR FTS measures the absorption spectra through a transmission/reflection path in the atmosphere or a test cell, or the spectra reflected or radiated from a solid surface. As a result, it often requires a calibrated source or reflector located on the opposite side of the volume to be interrogated. This is a viable approach in the laboratory or at fixed locations, but it is currently impractical for tactical and strategic applications.
Recent advances have implemented this technique with natural illumination as the source. This approach requires large training sets and extensive computation to remove background effects. Site contamination by the material of interest or by other chemical compounds require that training sets be collected at the test site prior to the introduction of real-time monitoring. Although promising, this approach is immature for tactical field deployment.
Raman Lidar transmits an intense pulse of monochromatic light, which stimulates the material in airborne gases, aerosols, liquids, and/or solids to radiate a spectrum of colors possessing wavelengths that are typically longer, but can be shorter, than the wavelength of the transmitted laser pulse. The offset of the wavelength of the radiated colors (i.e., the Raman spectrum) is a characteristic of the material of interest. However, while the offset is fixed, the wavelength of the transmitted laser light may be made variable. The transmitted light must be highly monochromatic, but the absolute wavelength of Raman Lidar is of less significance than in a DIAL system. As a result of the aforedescribed Raman process, the transmitter in a Raman Lidar may be considerably simpler than in a DIAL or DISC system.
Since all illuminated materials are exposed to the transmitted light simultaneously, all re-radiate their characteristic Raman spectra simultaneously. Thus, multiple species can be detected in parallel. The historical disadvantage of Raman Lidar is its lower sensitivity when it is compared to a DIAL system. This lower sensitivity is partially the result of the receiver architectures, which were designed for use in previous Raman Lidars. In addition, sensitivity is further reduced due to the relatively smaller Raman cross-section of the material as compared to the absorption cross sections observed in DIAL systems.
A Raman Lidar receiver commonly used is shown in FIG. 1A. Received light, via receiver element 8, is separated into its component wavelengths by a dispersive filter 6 (e.g., a grating or a prism). The separated light is imaged by imaging lens 4 onto a detector array 2.
The spectral resolution of the array 2 is determined by the characteristics of dispersive filter 6, the f-number of the imaging lens 4, and the size of the individual detectors. If the detector size is the limiting resolution parameter, the spectral range of the array 2 can be no greater than the resolution of a single detector multiplied by the number of elements in the array.
The very close spacing of some Raman lines often dictates high spectral resolution in many applications. The wide spectral separation of other lines simultaneously requires a large spectral range. Satisfying both requirements can necessitate a large number of detectors. Over 4000 detector elements (and in some cases 10,000) are not uncommon.
Charge coupled device (CCD) detector arrays are commonly used in Raman receivers since they permit a large number of detectors with a minimal number of electrical connections. Wiring 4000 individual detectors in parallel is impractical. Unfortunately, CCD detectors are not highly sensitive detectors. Furthermore, the overall sensitivity of the system is based upon its ability to detect the weakest (critical) line in the Raman spectra. An additional disadvantage is that the range resolution of this type of Raman Lidar is limited by the readout time of the detector array and typically not by the pulse length of the laser. Currently, a 4000 element CCD arrays can be readout in approximately 10 microseconds, which corresponds to a range resolution of about 1.5 kilometers. Thus, the speed at which data can be read via a CCD limits bandwidth, and therefore limits the range resolution.
Another Raman Lidar receiver commonly used is illustrated in FIG. 1B. Received light, via receiver element 8, is collected via collection optics 14 and imaged onto a spectral filter 12. The spectral filter will typically transmit only a single wavelength of light. The light which emerges from the spectral filter 12 is then imaged by lens 11 onto the single detector 10. In this regard, the spectral filter 12 is configured to enable only one wavelength of light to be passed through to the detector 10, and such a system typically operates effectively when a low concentration of an interferant exists in the light received by the receiver element 8.
The spectral filter 12 is typically designed to image the received light onto the single detector 10. In this regard, the spectral filter 12 is configured to only enable one predominant molecule corresponding to one wavelength of light to be passed through the detector 10, and such a system typically operates effectively when a low concentration of an interferant exists in the light received by the receiver element 8. Typically, these filters are made to observe atmospheric characteristics such as O2, O3, or N2. However, since only one, or a very small number of detectors are used, detectors having greater sensitivity and bandwidth than CCD detectors can be used. In this configuration signal as a function of distance can be recovered to improve range resolution.
The spectral filter 12 may be implemented using a variety of optical principles. For example, it may comprise a grating, which refers to an optical device consisting of a surface with many parallel grooves in it that disperses a beam of light into its component wavelength. Generally, the angle of the gratings determine the wavelength of light that is output from the grating, and the grating's resolution is determined by the number of lines in the grating. It may also comprise an interference filter, etalon, other devices or a combination of these techniques. Thus, with reference to FIG. 1B, the spectral filter 12 may be manipulated mechanically, electrically, and/or thermally, such that the filter is “tuned” to transmit a particular wavelength.
Although the system of FIG. 1B can have high sensitivity and bandwidth, there are various disadvantages. Most notably, much of the content of the received spectra is lost, because of all the wavelengths received in the spectra only one of them is transmitted and is available for detection by detector 10. Likewise, in order to subsequently obtain any information related to the wavelengths not previously detected, the spectral filter 12 must be tuned a plurality of times to detect the corresponding plurality of wavelengths. Further, for each wavelength that one desires to discern, a corresponding pulse must be initiated, resulting in a lengthy process.
As described herein, manipulation of the spectral filter 12 is usually performed by tuning to a particular wavelength, imaging the received light onto a detector 10, then retuning the spectral filter 12 to a different desired wavelength, thereby requiring that the process be repeated for each desired wavelength. Various monochromators provide this functionality by enabling a user to turn a knob on the monochromator to select various wavelengths, which adjust a grating within the monochromator.
Another disadvantage of DIAL and DISC systems which transmit multiple wavelengths, Raman systems which transmit multiple pulses and retune the receiver to observe multiple Raman spectral lines, and FTS systems which must collect signals for a significant period of time in order to obtain sufficient sensitivity is that the sample must not change during the measurement period. Changes in material composition during the measurement period can result in erroneous readings.
The second category for spectral measurement and identification is commonly referred to as point detection. The characteristic for this category is the specimen is in direct proximity to the instrument. For example, during atmospheric measurements when an operator is located in the same area as the sample and is exposed to the sample this is a point detection configuration. In contrast to point detection, stand off detection occurs when the operator is so remote from the sample that they are unaffected by its composition. Overall the same methods and techniques used for remote detection may also be used for point detection. In general, point detection systems may be more sensitive as they operate at shorter distances.